gcc 4.1.1: <tgmath.h>, decimal floating-point

gcc 4.1.1: <tgmath.h>, decimal floating-point

[Fred J. Tydeman]

This is my first build of gcc.

I downloaded gcc 4.1.1 onto a RedHat 8 Linux system on Intel Pentium 4
with its gcc 3.2.

I have
  srcdir = /misc/gcc-4.1.1/
  objdir = /FPCETS/gcc4.1.1/

I did
  cd objdir
  srcdir/configure --prefix=/FPCETS/gcc411 --enable-languages=c,c++
  make bootstrap
  make install

A compiler is built and mostly works.

However, <tgmath.h> is not usable.  In doing some research, I do NOT
find a <tgmath.h> in the 4.1.1 tree, but do find one in the 3.2 tree.
I assume that mixing a 4.1.1 compiler with a 3.2 header is a problem.
So, how do I build gcc 4.1.1 and force it to create all the C headers
in the 4.1.1 tree?

Also, I am writing tests of Decimal Floating-Point (DFP), but have
not found out how to have the compiler recognize _Decimal64 and the
other DFP types.  I assume that I am either building the compiler
wrong, or invoking the compiler wrong.  Any suggestions on how to
make DFP usable?

---
Fred J. Tydeman        Tydeman Consulting
tydeman@tybor.com      Testing, numerics, programming
+1 (775) 358-9748      Vice-chair of J11 (ANSI "C")
Sample C99+FPCE tests: http://www.tybor.com
Savers sleep well, investors eat well, spenders work forever.

gcc 4.1.1: <tgmath.h>, decimal floating-point

[Fred J. Tydeman]

This is my first build of gcc.

I downloaded gcc 4.1.1 onto a RedHat 8 Linux system on Intel Pentium 4
with its gcc 3.2.

I have
  srcdir = /misc/gcc-4.1.1/
  objdir = /FPCETS/gcc4.1.1/

I did
  cd objdir
  srcdir/configure --prefix=/FPCETS/gcc411 --enable-languages=c,c++
  make bootstrap
  make install

A compiler is built and mostly works.

However, <tgmath.h> is not usable.  In doing some research, I do NOT
find a <tgmath.h> in the 4.1.1 tree, but do find one in the 3.2 tree.
I assume that mixing a 4.1.1 compiler with a 3.2 header is a problem.
So, how do I build gcc 4.1.1 and force it to create all the C headers
in the 4.1.1 tree?

Also, I am writing tests of Decimal Floating-Point (DFP), but have
not found out how to have the compiler recognize _Decimal64 and the
other DFP types.  I assume that I am either building the compiler
wrong, or invoking the compiler wrong.  Any suggestions on how to
make DFP usable?

---
Fred J. Tydeman        Tydeman Consulting
tydeman@tybor.com      Testing, numerics, programming
+1 (775) 358-9748      Vice-chair of J11 (ANSI "C")
Sample C99+FPCE tests: http://www.tybor.com
Savers sleep well, investors eat well, spenders work forever.